Growing evidence supports the proposition that high dose therapy requiring hematopoietic support (transplantation) is a curative procedure for some patients with non-Hodgkin's lymphoma (NHL) who fail to achieve a remission or who relapse following conventional doses of chemotherapy. Furthermore, there is an indication that the source of hematopoietic support employed, autologous bone marrow (ABMT) versus autologous peripheral blood stem cells obtained by leukapheresis (PSCT) is also a determinant of outcome. In a non-randomized study, PSCT patients had a significantly better long- term disease free survival. In preliminary studies we have shown, using a sensitive culture technique to amplify lymphoma cells followed by Southern analysis to detect clonal populations, that histologically normal appearing bone marrow harvests were contaminated with tumor in about 30- 40% of patients. In contrast, using the same approach, blood stem cell harvests were contaminated in only 5-10% of patients. When the clinical outcome was analyzed for those patients who were recipients of tumor-free PSCT and ABMT they did much better than recipients of minimally contaminated marrow harvests. Therefore, it is our hypothesis that minimal disease in the harvest is a critical determinant of outcome in transplantation for lymphoma. In association with a proposed multicenter prospective randomized trial of ABMT versus PSCT for NHL patients with histologically uninvolved marrow, we will employ culture techniques as well as molecular biology, including Southern analysis and PCR analysis of appropriate sequences defined by the genetic rearrangement of the patient's primary tumor to detect lymphoma cells in the harvests. We will correlate the outcome and sensitivities of our methodological approaches to the clinical outcome of the patients in an attempt to confirm that minimal lymphoma in the harvest is a key determinant of outcome. We will evaluate if the extent of tumor contamination is a predictor of time to disease progression. These data should answer questions such as what is the optimum methodological approach for detection of minimal lymphoma and whether purging of marrow or selection of CD34 positive cells for transplantation is necessary or does the use of PSCT rather than AMBT suffice? This information should further improve the curative ability of transplantation for otherwise incurable NHL.